Measurement Invariance of Momentary Affect Across Time and Across Individuals from Adolescence to Old Age

This study tests the assumption that measures of momentary affect are invariant across individuals from different age groups and across time. Configural, metric, and scalar invariance of momentary affect will be examined for two measurement models: an affect-balance model, where positive and negative affect indicators load on the same latent factor; and a positive-and-negative-affect (PA-NA) model with two latent factors. Analyses will be based on existing data of a multi-wave experience sampling study: the Multi‐Method Ambulatory Assessment (MMAA) Study. We will use three waves collected in the years 2007 (N = 378 participants), 2010 (N = 398), and 2013 (N = 365). At each wave, the experience sampling phase consisted of six daily assessments on nine days throughout three weeks (M = 54 assessments per participant). We will use cross-classified factor analysis for intensive longitudinal data to examine measurement invariance of the affect-balance and the PA-NA models. Cross-classified factor analysis is based on Bayesian estimation. The analysis reveals to what extent the within-person measurement model parameters (loadings for metric invariance, additionally intercepts for scalar invariance) show variation across individuals and across measurement occasions. unknown other